Incomptatible versions of GCC and CUDA in yml file

[DylanWangWQF]

Hi @SyedShaQutub
In the yml file, you specify the version of GCC (e.g., gcc_linux-64) as 11.2.0 and CUDA as 11.3, however, it outputs the following mismatch error:
The current installed version of /home/qifan/anaconda3/envs/pytorchALFI/bin/x86_64-conda-linux-gnu-c++ (11.2.0) is greater than the maximum required version by CUDA 11.3 (10.0.0). Please make sure to use an adequate version of /home/qifan/anaconda3/envs/pytorchALFI/bin/x86_64-conda-linux-gnu-c++ (>=5.0.0, <=10.0.0).

The cause of this iscudatoolkit? I guess some versions in condaenv.yml should be updated?

Some of my testbed:

conda 24.1.2

//  cudatookit 11.3 is installed to match the `yml` file
nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2021 NVIDIA Corporation
Built on Mon_May__3_19:15:13_PDT_2021
Cuda compilation tools, release 11.3, V11.3.109
Build cuda_11.3.r11.3/compiler.29920130_0

[SyedShaQutub]

Hi @DylanWangWQF,
Thanks for trying out our repo. The yml file contains some issues, especially the unnecessary packages. Regarding the GCC version mismatch, you could update the code to the latest version or install the combined gcc to match it with Cuda.
There is also another solution; please refer to this page - https://stackoverflow.com/questions/6622454/cuda-incompatible-with-my-gcc-version.

Let me know if it works

[DylanWangWQF]

@SyedShaQutub
Thanks for your suggestion! I will try later since my GPU machine doesn't work now.

you could update the code to the latest version

It's the latest. I just cloned the repo several days ago. So I think anyone sets up according to yml file recently, he will get the same error.

install the combined gcc to match it with Cuda.

I will try to ln -s GCC link in conda env and see if it works.